While troubleshooting a mail server failure today (lots of local 'mail transport unavailable' errors in /var/log/zimbra.log), I discovered that the /opt/zimbra/log/mailbox.log had grown to over 2 GB (usual daily size is around 50MB)! Most of the log was filled with errors like the following. I see this error appear for multiple users:
2009-04-13 08:43:36,546 INFO [LmtpServer-37415] [email@example.com;mid=80;] index - MILucene23Index at com.zimbra.cs.index.Z23FSDirectory@/opt/zimbra/index/0/80/index/0LRU empty and all slots reserved...retrying
I was able to get the mail service operational again by restarting zimbra (zmcontrol stop; zmcontrol start). Can anyone help me understand what all these index-related errors might mean?
Release 5.0.13_GA_2791.RHEL5_64_20090206112057 CentOS5_64 NETWORK edition